Multi-resolution convolutional neural networks for fully automated segmentation of acutely injured lungs in multiple species
Multi-resolution convolutional neural networks for fully automated segmentation of acutely injured lungs in multiple species
复制标题
DOI：
10.1016/j.media.2019.101592
复制
发表时间：
2020-02-01
影响因子：
10.9
通讯作者：
Reinhardt, Joseph M.
中科院分区：
文献类型：
--
作者：
Gerard, Sarah E.;Herrmann, Jacob;Reinhardt, Joseph M.
急性呼吸窘迫综合征（ARDS）肺的分割是一项具有挑战性的任务，由于弥漫性混浊的依赖区域，导致很少或没有对比度的肺边界。对于严重损伤的肺的分割，局部强度和纹理信息以及全局上下文信息是用于一致地包含肺内结构的重要因素。在这项研究中，我们提出了一个深度学习框架，该框架使用一种新型的多分辨率卷积神经网络（ConvNet）来自动分割具有类似于ARDS损伤模型的多个哺乳动物物种的肺部。多分辨率模型通过使用低分辨率到高分辨率网络的级联，消除了在高分辨率和全局上下文之间进行权衡的需要。迁移学习用于适应有限数量的训练数据集。该模型最初在人类CT图像上进行预训练，随后在具有类似于ARDS的肺损伤的犬、猪和绵羊CT图像上进行微调。将多分辨率模型与单独的高分辨率和低分辨率网络进行比较。多分辨率模型优于低分辨率和高分辨率模型，动物数据集（N = 287）的总体平均Jacaard指数分别为0.963 +/-0.025和0.919 +/- 0.027和0.950 +/- 0.036。多分辨率模型实现了0.438 +/- 0.315 mm的整体平均对称表面距离，而低分辨率和高分辨率模型分别为0.971 +/- 0.368 mm和0.657 +/- 0.519 mm。我们的结论是，多分辨率模型产生准确的分割在严重受伤的肺，这是由于包括本地和全球的功能。(C)2019 Elsevier B. V.版权所有。
Segmentation of lungs with acute respiratory distress syndrome (ARDS) is a challenging task due to diffuse opacification in dependent regions which results in little to no contrast at the lung boundary. For segmentation of severely injured lungs, local intensity and texture information, as well as global contextual information, are important factors for consistent inclusion of intrapulmonary structures. In this study, we propose a deep learning framework which uses a novel multi-resolution convolutional neural network (ConvNet) for automated segmentation of lungs in multiple mammalian species with injury models similar to ARDS. The multi-resolution model eliminates the need to tradeoffbetween high-resolution and global context by using a cascade of low-resolution to high-resolution networks. Transfer learning is used to accommodate the limited number of training datasets. The model was initially pre-trained on human CT images, and subsequently fine-tuned on canine, porcine, and ovine CT images with lung injuries similar to ARDS. The multi-resolution model was compared to both high-resolution and low-resolution networks alone. The multi-resolution model outperformed both the low- and high-resolution models, achieving an overall mean Jacaard index of 0.963 +/- 0.025 compared to 0.919 +/- 0.027 and 0.950 +/- 0.036, respectively, for the animal dataset (N = 287). The multi-resolution model achieves an overall average symmetric surface distance of 0.438 +/- 0.315 mm, compared to 0.971 +/- 0.368 mm and 0.657 +/- 0.519 mm for the low-resolution and high-resolution models, respectively. We conclude that the multi-resolution model produces accurate segmentations in severely injured lungs, which is attributed to the inclusion of both local and global features. (C) 2019 Elsevier B.V. All rights reserved.